STABLE ISOTOPE BASED PALEOALTIMETRY AND PALEOHYPSOMETRY

Paleoaltimetry and paleohypsometry are crucial data in the evaluation of large-scale tectonic models of the evolution of orogenic systems. Until very recently geologists have had virtually no reliable means of directly determining paleoaltitude or the more important paleohypsometry of any mountain belt. Rowley et al. (2001- EPSL 188, 253-268) presented a parcel-based model that combines atmospheric thermodynamics of water vapor condensation with isotopic fractionation of oxygen and hydrogen isotopes to predict the relationship between elevation and isotopic composition in orographically forced precipitation systems for latitudes equatorward of about 40°. Good correlation of predictions and observations using the Global Network of Isotopes in Precipitation data (from IAEA) and with a limited collection of water samples from the Himalayas and southern Tibet provided confidence of the robustness of the model. Surface waters, including lakes, rivers and soil waters will be the source for virtually all geological samples used for paleoaltimetry. Thus it is most appropriate to make comparisons between model predictions and observations of these types of waters. The comparisons of Rowley et al (op. cit.) are extended using a significantly larger dataset of both published and recently collected modern waters from the Himalayas, southern and central Tibet, and from the network of rivers draining the Himalayas into the Indo-Gangetic basin. The isotopic composition of surface water is determined by the hypsometry of the drainage basin above the sampling site weighted by the amount of precipitation as a function of elevation. Thus each river will have its own variation of isotopic composition with elevation. The Rowley et al. model is used to predict both the hypsometric mean elevation and condensation weighted hypsometric mean elevation of sampled drainage basins and compare these with observation. There is a good correlation between model-based predictions and observations using modern waters. Application of this model to samples of paleolakes and paleosols in the Himalayas and Tibet for samples ranging from 10 to 40 million years old provides for the first time a picture of the evolution of Himalayas and Tibet since collision began in the Middle Eocene